How-to guide: How to comply with legal requirements relating to selection criteria during a redundancy process (UK)

Introduction

This how-to guide assists in-house counsel, private practice lawyers and human resources professionals working in England, Wales and Scotland (GB) to achieve compliance with the requirements for using selection criteria during a redundancy exercise. This guide does not apply to Northern Ireland, where employment law is devolved to the Northern Ireland Assembly.

Redundancy is a potentially fair reason for dismissal under section 98(2)(c) of the Employment Rights Act 1996 (ERA 1996). It covers situations where the employer has a reduced need for employees to carry out work of a particular kind, or where a workplace is closing.

This guide covers:

Designing and applying objective selection criteria
Scoring methodologies and consistency
Ensuring transparency and auditability
Common pitfalls and legal risks

This guide should be read in conjunction with Checklist: Conducting a redundancy exercise, which provides an overview of the key stages of a redundancy process, and How-to guide: How to establish a fair selection pool during a redundancy process.

Employers should be aware that the Employment Rights Bill 2024, introduced on 10 October 2024, aims to significantly enhance worker protections in the UK. Key provisions include granting day-one rights for unfair dismissal, paternity, parental, and bereavement leave, changes to sick pay and flexible working rights, protections against zero-hours contracts and restrictions on the use of ‘fire and rehire’ practices. New legislation will be introduced throughout 2025 and 2026. Further information can be found at UK Parliament: Employment Rights Bill and Employment Rights Bill: factsheets.

Section 1– Designing and applying objective selection criteria

Once an employer has established the pool of employees from which redundancies will be made, employees must be selected from that pool using selection criteria. For information about redundancy selection pools, see How-to guide: How to establish a fair selection pool during a redundancy process.

In workplaces with recognised trade unions or elected employee representatives, the selection criteria should, wherever possible, be the subject of meaningful consultation with those representatives. This not only supports compliance with collective consultation duties (see Quick view: Collective redundancy consultation), but can also improve buy-in from the workforce. In addition, each affected employee should have an opportunity to discuss how the proposed criteria will be applied to them in practice, including the evidence relied upon. Such consultation helps identify and address any inaccuracies or unfair assumptions before final decisions are made.

Selection criteria are a key part of any fair redundancy process. If the selection pool is reasonable but the criteria are flawed, the dismissal may still be unfair. The criteria used must be lawful, objective, and fairly applied. This section focuses on what that means in practice.

1.1 What makes selection criteria lawful and objective?

Lawful criteria are those that do not directly or indirectly discriminate against protected groups. Under the Equality Act 2010 (EqA 2010), employers must avoid selection criteria that disadvantage people based on age, sex, race, disability, maternity, or other protected characteristics. For further information about discrimination generally, see How-to guide: Overview of workplace discrimination and harassment law and Quick view: Protected characteristics under the Equality Act 2010.

Objective criteria are those that can be measured or assessed in a way that does not depend on personal opinion. For example, disciplinary records, attendance, qualifications, and formal performance reviews are all potentially objective, so long as the evidence behind them is accurate and applied consistently.

The safest approach is to use multiple objective criteria, each with clear definitions and scoring guidance. For example, if performance is being assessed, the employer might use appraisal scores from the last two years, measured against set competencies. If attendance is used, it should exclude absences related to disability, pregnancy, or maternity. This avoids unlawful discrimination and improves fairness.

The criteria must also relate to the job. If the business is reducing customer service roles, it makes sense to assess customer-handling skills, communication, and complaint resolution — not unrelated factors like finance qualifications or IT training.

In-house counsel should be involved in reviewing criteria before they are used. Ask whether each criterion can be measured. Ask what evidence will be relied on. If a criterion cannot be explained clearly or scored in a consistent way, it probably should not be used.

1.2 Non-discriminatory selection criteria

In order to be lawful, selection criteria must avoid both direct and indirect discrimination under the EqA 2010. Employers need to consider carefully how each criterion might affect employees with protected characteristics. Even if the criteria seem even-handed, the effect can still be discriminatory.

Indirect discrimination occurs when an employer applies a provision, criterion or practice (PCP) that appears neutral but puts people with a protected characteristic at a disadvantage, without good reason. This is a key risk in redundancy selection. Employers often use criteria like attendance, flexibility, or shift patterns without realising how these may impact certain groups more than others.

Under section 19 of the EqA 2010, a PCP is indirectly discriminatory if:

it is applied to everyone;
it puts people with a particular protected characteristic at a disadvantage;
it puts an individual with a particular protected characteristic at that disadvantage; and
the employer cannot justify it as a proportionate means of achieving a legitimate aim.

If a scoring matrix penalises people for having more sick days, that may indirectly discriminate against disabled employees. People with long-term health conditions are more likely to need time off. Even if the scoring is not intended to discriminate, the effect is what matters. To avoid this, employers must exclude or adjust for disability-related absence when using attendance scores.

Example – maternity leave

Maternity leave is an area where mistakes happen. Employers sometimes treat maternity leave as neutral time — not scoring at all — or mark employees down for not having recent appraisals. This can be deemed as indirect sex discrimination. A fair process should ensure that maternity absence does not disadvantage the employee. That may mean using the most recent available data or adjusting the criteria so that those on leave are not penalised.

Example – flexibility

Flexibility and availability for extra shifts can be a risk area. Criteria based on willingness to work late, travel, or work weekends may disadvantage women more than men, as women are statistically more likely to have caring responsibilities for children. These may be deemed as indirectly discriminatory on the grounds of sex unless they are essential for the role and the employer can show that there is no less discriminatory way to meet the business need.

Example – religious observance

The protected characteristic of religion and belief can be affected. A scoring system that rewards weekend availability might disadvantage employees whose religious observance prevents them from working at certain times.

Example – length of service (LIFO)

Length of service, or last in first out (LIFO), is often seen as simple and neutral, but it disadvantages younger workers and can lead to indirect age discrimination. While it is not unlawful to use LIFO, it should never be the sole or primary criterion unless the employer has a clear and well-documented justification. Even then, tribunals are cautious. Most employers now avoid using LIFO at all or combine it with other factors.

The law allows indirect discrimination if the employer can justify it. The employer must show that the aim is legitimate — for example, maintaining service levels — and that the means used are proportionate. That includes showing that less discriminatory alternatives were considered and rejected for sound reasons.

In practice, employers often face claims not because the criteria were unlawful on their face, but because the employer failed to consider who would be affected. In-house counsel can play a key role by stress-testing proposed criteria against workforce data, past HR issues, or known employee characteristics.

In-house counsel should review proposed criteria with these risks in mind. Ask whether any protected group is likely to be disadvantaged, even unintentionally. If so, consider whether the criterion can be changed, qualified, or adjusted. Being able to show that equality risks were considered in advance is one of the best protections an employer can have if the process is challenged.

1.3 Criteria based on performance and skills

When choosing selection criteria, employers often focus on performance, skills, and attendance. These are all valid — but only if they are applied in a way that is evidence-based, consistent, and legally sound. Each carries different risks and needs to be handled carefully to ensure fairness and compliance.

1.3.1 Performance

Performance is a legitimate criterion and employers are entitled to retain the best performers. But the key is to base this on objective records, not informal impressions. Appraisal scores, performance ratings, achievement of measurable targets, or structured competency assessments are safer than general views about who is ‘strong’ or ‘weak’. If line managers are asked to score performance directly, they must use consistent criteria and explain how scores were reached. Subjective scoring without clear evidence is one of the most common reasons redundancy dismissals are challenged.

Where recent appraisals do not exist, employers should not guess or create scores on the spot. If no objective record is available, it may be better to use another criterion. In-house counsel should ask what evidence is being relied on for each score — if the answer is vague or inconsistent across the group, the criterion should be reviewed.

1.3.2 Skills

Skills are also a fair basis for selection. The employer may want to retain people with particular technical abilities, qualifications, or experience. But again, there must be evidence. Job descriptions, qualification records, training logs, and skills matrices can all help. If the employer values adaptability or multi-skilled employees, it must be clear how that has been assessed and why it matters to the future business structure.

One risk here is over-reliance on informal knowledge. If one manager scores someone highly for having a certain skill, but another manager overlooks that same skill in someone else, the inconsistency can undermine the process. It helps to define each skill or competency clearly in advance, so that everyone being scored is assessed against the same benchmark.

1.3.3 Avoiding vague criteria

Employers should also be cautious around criteria like ‘team fit’, ‘commitment’, or ‘attitude’. These are vague and open to interpretation. They may sound fair, but they are difficult to apply consistently and are often used to disguise personal preference. If such criteria are used, the employer must be able to show how they were defined and assessed fairly. In many cases, it is safer to leave them out.

Section 2 – Scoring methodologies and consistency

Once the selection criteria are agreed, the next step is to apply them consistently. This usually involves scoring each employee in the pool against each criterion. Scoring helps ensure that decisions are structured and based on evidence. But a scoring system is only as fair as the process behind it. Inconsistency, lack of evidence, or poorly designed scoring ranges can quickly undermine the process.

2.1 Selection matrices and weighting

A selection matrix is a tool for applying redundancy criteria in a structured way. It lists each employee in the pool, sets out the chosen criteria, and records the score for each one. The aim is to create a clear and consistent record of how the selection decision was made. Done well, it helps show the process was objective and fair. Done badly, it creates confusion and legal risk.

Most matrices use a scoring range — usually 1 to 5 — for each criterion. Each employee is scored and the totals are added to give a final result. Higher scores indicate stronger performance against the chosen measures, and employees with the lowest totals are at risk of redundancy. The system is simple, but it only works if the scoring is clear, consistent, and based on evidence.

2.1.1 Scoring guidance

Scoring definitions should be written down and shared with all scorers. What counts as a ‘3’ or a ‘5’ must be consistent across the board. For example, if the employer is scoring on performance, what specifically justifies a top score? Is it exceeding all targets? Consistently positive feedback? Without clear definitions, scoring becomes subjective — and once that happens, the whole process is open to challenge.

2.1.2 Weighting

Weighting adds another layer. Some employers treat all criteria equally. Others assign more importance to certain areas — for example, giving double weight to performance and less to attendance or disciplinary record. This can reflect business priorities, especially where the organisation needs to retain specific skills or capabilities.

There is no rule against weighting, but it needs to be thought through. First, the rationale should be recorded. Why is one area more important than another? Can the business justify that decision if asked in a tribunal? Second, the weighting must be applied consistently. It cannot change mid-way through the process, and all employees in the pool must be scored using the same version of the matrix.

Problems arise when weighting is used to tip the balance in a particular direction. For example, increasing the weight of a criterion that favours a particular group, or changing the weighting after scoring has started, can undermine the process. Tribunals will expect to see that the weighting was planned in advance and applied impartially.

Another issue is overcomplicating the matrix. Some employers try to include 10 or more criteria, each with its own weighting, and end up with a system no one really understands. In practice, a simpler matrix with three or four clear, relevant criteria — properly weighted and supported by evidence — is far safer. The goal is clarity, not complexity.

In-house counsel should check that the matrix has been designed in a way that reflects the business need, complies with legal obligations, and can be explained clearly if challenged. They should also check that it has been applied consistently across the pool, with proper documentation at each stage.

2.2 Calibration and review processes

Once scoring begins, consistency becomes critical. Even with a clear matrix and defined criteria, different managers can apply scores in very different ways. That is where calibration and review processes come in. They help ensure fairness across the pool and reduce the risk of subjective or inconsistent decisions.

Calibration means sense-checking the scores given by different managers. If several line managers are involved in scoring, each may interpret the same criterion slightly differently. One may routinely give 4s where another gives 2s for similar performance. Without calibration, employees are not being treated equally — even if the scoring method is technically the same.

A simple calibration meeting can make a big difference. Bring the relevant managers and HR together once initial scoring is complete. Review the spread of scores. Are there any outliers? Are similar employees being scored very differently? If so, ask why. This is not about forcing everyone to score the same — it is about checking that the differences are based on evidence, not inconsistency or unconscious bias.

Calibration is particularly important for subjective areas like performance or attitude. These categories should already be supported by specific examples or appraisal data. But if scoring still seems inconsistent, it may need to be reviewed. The employer must be able to show that any difference in scores reflects actual differences in performance, not differences in management style or personal opinion.

Review processes also matter at the organisational level. HR and in-house counsel should review the overall scoring outcomes. That includes checking that:

the criteria have been applied consistently;
the weighting (if used) has been handled correctly; and
protected characteristics have not inadvertently influenced scores.

In some cases, it is worth anonymising results during the review to reduce bias. For example, reviewing scores without names can help focus attention on the evidence, not the individual.

Another point to check is how absences, flexible working, or recent changes in duties have been treated. Employees returning from maternity leave or long-term sick leave often have limited data available. The employer must ensure that this hasn’t resulted in lower scores or unfair assumptions. If necessary, alternative evidence should be used, or adjustments made.

It is also important to ensure that all records are complete. Each score should be supported by some form of evidence — whether it is an appraisal, a training record, or absence data. If that evidence is missing or inconsistent, the score should be reviewed and, if needed, revised before the process moves forward.

Section 3 – Ensuring transparency and auditability

A redundancy process does not just need to be fair – the employer must be able to prove it was fair. That is where transparency and auditability come in. Every key decision, particularly around scoring and selection, should be recorded in a way that can be explained later. If the matter ends up in a tribunal, having a clear, well-documented process makes all the difference.

3.1 Documenting how scores were reached

Employers must be able to show how each redundancy score was reached. It is not enough to produce a spreadsheet with numbers — there needs to be a clear link between the score and the evidence behind it. This applies to every criterion used in the selection matrix. Without that link, the process risks looking arbitrary, or at worst, biased.

Each score should be supported by some form of record. For performance, that might include recent appraisals, performance improvement plans, or objective data like sales figures or KPIs (key performance indicators). For attendance, it should be HR records showing days off, with clear notes explaining whether disability-related or maternity absences were removed. For skills, training logs, qualification records or examples of relevant work can all help justify the score.

This evidence does not need to be included in the matrix itself, but it must be available and capable of being shared if the employee asks for it or if the matter ends up in tribunal. A good approach is to keep a short-written explanation alongside each score — a sentence or two explaining why the employee received a ‘3’ or a ‘5’ under that heading. This makes it easier to explain the decision later and helps ensure consistency between scorers.

Where multiple managers are involved in scoring, consistency becomes even more important. Two employees doing the same job shouldn’t get different scores just because their managers apply the criteria differently. That is why internal moderation or calibration is essential, but so is documenting the reasoning. If one employee gets a 4 for performance and another doing the same work gets a 2, the business must be able to show why.

It is also important to record how any adjustments were made. For example, if a manager originally gave a lower score but revised it after discussing the evidence with HR, that change should be documented with a note of why it was made. Without this, the employer risks the tribunal viewing the process as unstructured or influenced by pressure.

If the employee requests their scores — or raises concerns — the explanation should already be written down. Trying to reconstruct reasoning after the fact rarely works well. Memories fade, and tribunals are generally sceptical of retrospective explanations that are not supported by contemporaneous notes.

In-house counsel should review a sample of scoring records before the process is finalised. Ask whether the reasoning is clear, whether the score matches the evidence, and whether the same standard has been applied across the board. If it’s not clear how a score was reached, it may need to be revisited.

3.2 Internal moderation and consistency

Once scoring is complete, the employer needs to check that it stands up to scrutiny through internal moderation.

Moderation is not a formal legal requirement, but it is widely accepted as good practice. It usually takes the form of a meeting where HR, line managers, and sometimes the in-house legal team review the scores and check for inconsistencies. The aim is not to change outcomes arbitrarily, but to make sure the process has been applied fairly and that decisions are supported by evidence.

One of the main issues that comes up during moderation is score inflation or deflation. Some managers score generously, others more harshly. If two employees with near-identical records receive very different scores, that is a problem. It suggests the scoring process was not applied consistently and that undermines fairness. During moderation, these discrepancies can be identified and either justified or corrected. The decision to change a score must be recorded, with a short explanation of why the change was made.

Another issue is inconsistency in the use of evidence. One manager might rely heavily on documented appraisals, while another bases scores on memory or informal knowledge. All scorers should be using the same standard of evidence.

Moderation helps flag situations where one score is well-supported and another is not. In some cases, moderation may also reveal gaps in the scoring process. For example, a manager may have given a score without recording the reason. Or they may have misunderstood the scoring scale. These are opportunities to fix the issue before final decisions are communicated.

This review process can also help identify indirect discrimination risks. If employees with protected characteristics are consistently scoring lower in certain areas, the employer should pause and check why. Is the criterion itself problematic? Has a manager made assumptions? Even if the outcome doesn’t change, showing that these questions were asked and considered can be vital if the process is challenged.

The outcome of moderation should be recorded. This does not need to be complex — a short note explaining who attended, what was reviewed, and whether any changes were made is enough. Any amendments to scores should also be documented in the individual record, with a brief explanation. This transparency shows the employer has taken care to apply the process fairly.

In-house counsel should be involved in or have oversight of moderation where possible. Their role is to spot legal risks, flag inconsistencies, and ensure that the scoring process aligns with the employer’s obligations under employment and equality law.

Section 4 – Common pitfalls and legal risks

Even where a redundancy is genuinely needed, employers often make avoidable mistakes when defining the pool or applying selection criteria. These errors do not always stem from bad faith — many are the result of rushed decisions, inconsistent processes, or assumptions that go unchallenged.

This section highlights the most common problems with selection criteria that lead to legal risk. Some relate to how decisions are made, others to how they are recorded or explained. What they have in common is that each can undermine an otherwise lawful process — and turn a fair dismissal into an unfair one.

Clear documentation, consistent application of criteria, and a willingness to challenge assumptions are usually enough to avoid most of these traps.

Common mistakes by employers when dealing with selection criteria include the following.

4.1 Applying selection criteria inconsistently

One of the quickest ways to undermine a redundancy process is to apply the selection criteria inconsistently. Even where the criteria are lawful and sensible, inconsistent application can make the outcome unfair. This is often the result of poor preparation, unclear definitions, or a lack of coordination between managers.

The most common problem is variation between scorers. This may be due to:

different managers scoring staff using their own judgement; and
inconsistent use of evidence (eg, some managers basing their scores on documented appraisals, while others rely on memory or informal feedback).

Problems also arise when the scoring system changes mid-process. This might be due to a change to the weighting of criteria, a redefinition of the scoring scale, or introducing a new criterion entirely. If this happens after scores have already been given — especially if it changes the outcome — the tribunal is likely to take a dim view. Even if the intention was to ‘fix’ the process, the employer may simply have created another flaw.

To avoid these problems, scoring guidance should be prepared in advance. Managers must be briefed on how to use it and reminded to score based on evidence, not personal opinion. A short calibration or moderation meeting can help check for inconsistencies before scores are finalised. Any changes to scores or adjustments made during this process should be recorded, with a clear explanation.

4.2 Overreliance on manager knowledge

In a redundancy process, relying too heavily on what managers ‘know’ — without supporting evidence — is a common and risky mistake. Tribunals need to see how decisions were made, backed up by supporting evidence comprising more than just instinct or memory alone.

This issue often arises in smaller teams, where line managers may have worked with the same staff for years. It is easy to assume that familiarity means fairness — but it also makes unconscious bias more likely.

Tribunals will look closely at how the employer reached its conclusions. If the selection process was based mostly on informal impressions — without documented examples or clear scoring standards — the dismissal may be found unfair, even if the redundancy situation itself was genuine.

4.3 Discriminatory scoring criteria

Selection criteria that appear neutral can still lead to unlawful discrimination, as outlined at section 1.2 above.

Tribunals do not just look at whether discrimination happened — they ask whether the employer considered the risk of it happening and took steps to avoid it.

In-house counsel should review the proposed criteria before the process begins and consider whether any criteria disadvantage any group with a protected characteristic. If so, can it be justified as a proportionate means of achieving a legitimate aim? If not, it should be removed or adjusted.

4.4 Use of redundancy to manage conduct or performance

Redundancy is not a shortcut for avoiding proper processes around misconduct, poor performance, or relationship breakdowns. Using redundancy to remove an employee for reasons unrelated to role reduction is one of the clearest ways to create legal risk. Tribunals are alert to this and will test whether the redundancy was genuine or a cover for another motive.

Employers sometimes fall into this trap when a redundancy situation coincides with concerns about an individual’s conduct or performance. Rather than go through a capability or disciplinary process, the employer includes the person in a small pool or singles them out for redundancy. Even if the business case is real, the selection decision must still be based on fair and objective criteria. If the employee appears to have been targeted because they were already seen as a problem, that risks a finding of unfair dismissal.

It is also common for employers to use vague or subjective criteria — like ‘attitude’, ‘reliability’, or ‘team fit’ — to reflect performance concerns that haven’t been formally addressed. This is risky. If the employer wants to factor performance into scoring, it must rely on actual evidence: recent appraisals, documented improvement plans, or clear examples of underperformance that have been raised at the time. If no such evidence exists, or if the employee was told previously that their performance was satisfactory, the employer may struggle to justify the score.

Using redundancy in place of capability or conduct processes also cuts out the protections that normally apply — such as the right to improvement time, support measures, or disciplinary warnings. Tribunals may conclude that the redundancy process was a sham or that the dismissal was substantively unfair.

There are also risks beyond unfair dismissal. If the employee has raised grievances, blown the whistle, or complained of discrimination, and then finds themselves selected for redundancy, they may claim victimisation or automatic unfair dismissal. In those cases, the employer’s motive is examined closely. If the evidence shows that redundancy was used as a convenient label, the claim may succeed even where there was a genuine business change.

In-house counsel should always ask whether any selection decision could be perceived as driven by performance or conduct concerns — and whether those issues have been properly addressed in their own right. If not, they should be kept separate from redundancy decisions.

4.5 Absence of documentation or transparency

Even when the decision to make redundancies is justified and the process appears fair, employers often come unstuck because they cannot show what happened or how decisions were made. Lack of documentation and transparency is one of the most common and avoidable reasons redundancy dismissals are challenged successfully.

In-house counsel should always check whether documentation exists to support each stage of a redundancy process: the business case, the pool, the scoring matrix, the evidence behind the scores, and consultation outcomes. If the only records are a spreadsheet and a termination letter, that is not enough to defend a tribunal claim. Ultimately, a fair process is one the employer can explain and prove.